ANOVA after logistic regression?

#1
I have fire incidences data with variables - seasons (summer & winter), time taken by fire truck to arrive (Early (<30 mins) and Late (>30 mins)), area (area code), small fire incident (1=Yes, 0=No), and big fire incident (1=Yes, 0=No). It has multiple observations recorded from multiple locations over the last year. I have a question in my mind:
"Whether the frequency of small fire converting to big fire incidence increase in summer while adjusting for the arrival time of arrival truck and area?"
Assuming that the probability of a small fire converting to big is 50% (Null hypothesis). Is there a way by which I can make a better model to analyze this type of data (probably through ANOVA on log-odds)?
I will be grateful if someone could provide me with a solution to this question. If it helps, I can provide a subset of the data.
 

Karabiner

TS Contributor
#3
small fire incident (1=Yes, 0=No), and big fire incident (1=Yes, 0=No).
So, essentially you want to predict "big fire yes/no"?
(probably through ANOVA on log-odds)?[/quoote]
I am not sure what you mean. The dependent variable is binary (big yes/no), therefore a binary logistic
regression seems appropriate. For ANOVA, you need an interval scaled dependent variable.

With kind regards

Karabiner